PCR optimizatio

Parkinson Lab parkinsonLab at bioscience.utah.edu
Mon Jan 24 00:32:47 EST 1994

In article <2hn69dINNalk at newsstand.cit.cornell.edu>, douglas l feinstein
<dlfeins at cumc.cornell.edu> wrote:

> I have just started on the bio.net news and came across a number of PCR
> related topics. One problem we have been running into concerns the
> presence of a large amount of diffuse background bands, running from as
> high as 2kB down to 100 bp, following 35 cycles of amplification. Our
> templates are a single band  of about 400 bp that has been purified
> through 2% Nusieve LMP. We do not purify the DNA<, simply melt the
> agarose and dilute it 1:1000 to use as template. On our gels, we do see
> the expected size product, but there is this awful background problem. In
> the absence of added template, there is not background, so our buffers,
> oligos, etc.. are clean. If anyone has ideas we sure would appreciate it.
> 	Regarding PCR machines, I have just purchased 3 Hybaid Thermocyclers,
> their original machines, after checking out the new ericcomps (their
> peltier system), a Coy, and a Mycycler. The Hybaids are really easy to
> use, have tube control, ramping ability, and our relatively cheap. We
> paid about $2700 for each, and got a couple of racks of aerosol tips
> thrown in. Their one disadvantage is the air cooling, limiting the low
> temp to ambient. However, we always leave our samples sitting at 30o C
> ON, with no problems. They also suggested we could put one machine in the
> cold room and could then program it down to 4oC.
> 	We are also interested in competitive PCR RT, in particular the methods
> people are using to prepare competitive templates. We'd be happy to share
> some ideas.
> Doug Feinstein
> Dept Neurobiology
> Cornell Univ Med College.
> dlfeins at cumc.cornell.edu

Dilute your starter DNA E-2 to E-3. This will eliminate problems of
contamination (agarose, Mg++, etc). Don't worry about the dilution.
Remember, the whole point of PCR is geometric amplification. If you dilute
E-2, for example, you get back the stock a little after 6 cycles. 

Make 3 reactions containing 1, 2 and 3 mM MgCl2. Divide each into 3
aliquots and run them at annealing temperatures of 40, 50 and 60C. This
generates a 3 x 3 matrix of Mg++ concentration vs. annealing temperature.
Inevitably, at least one of the combinations proves optimal.

Load the 9 reactions on a gel as follows to display them in order of
increasing stringency:


where H,M & L mean 3, 2 & 1 mM MgCl2.

Most of these techniques were suggested by Randy Rasmussen of Idaho
Technology, Inc. I have used this protocol hundreds of times. It *always*
works, as long as reagents are good and the primers are even slightly
balanced. It adds some overhead with traditional machines, but is well
worth it. I use Idaho Tech air thermal cyclers, which add only a couple of
extra hours for optimization. I find this inconsequential when weighed
against the benefits.

I am associated with Idaho Technology only as a satisfied user.

Eric Kofoid.

More information about the Methods mailing list